Eleven Months on Sotorasib, and a Pleural Effusion Nobody Has Tested
A single patient progressing after eleven months on sotorasib. The real disagreement isn't which regimen is better in the abstract — it's whether committing to either one before an already-collected pleural fluid sample gets tested is responsible.
W.T., 67, spent 34 years driving overnight freight routes before retiring, and jokes that his lungs “logged more miles than the truck did.” He quit smoking eight years ago, after 40 pack-years, following a bad case of pneumonia that finally scared him into it. Two years ago he was diagnosed with stage IV lung adenocarcinoma carrying a KRAS G12C mutation, found on his initial biopsy. He had four cycles of first-line carboplatin, pemetrexed, and pembrolizumab and progressed after roughly eight months; only then did he start sotorasib, whose accelerated approval (on CodeBreaK 100, and confirmed against docetaxel in the phase III CodeBreaK 200) covers KRAS G12C-mutant NSCLC only after at least one prior systemic therapy. Sotorasib produced a partial response within weeks, along with real improvement in a chronic cough he'd stopped mentioning to anyone because he'd assumed it was permanent. Surveillance imaging three weeks ago told a different story: a new small right pleural effusion and interval growth of the primary right upper lobe mass — objective progression, not measurement noise. Thoracentesis was performed for symptom relief, and cytology confirmed malignant cells, but the fluid itself was never sent for molecular testing — an available specimen sitting unexamined for the actual reason his cancer started growing again.
Acquired resistance to KRAS G12C inhibitors doesn't arise through one route. Some tumors develop a secondary mutation that blocks the drug from binding at all; others reactivate the same growth pathway through an entirely different receptor, most often MET amplification; a smaller number transform histologically into small-cell carcinoma, a change that would make several of the drugs under discussion here essentially irrelevant. Each possibility argues for a different next step, and none has actually been ruled in or out for him yet. He has otherwise done reasonably well on treatment — his COPD, a real legacy of those forty pack-years, has stayed stable on his usual inhalers throughout, and he's kept driving to VFW meetings most weeks despite the fatigue. What makes the untested fluid more than a procedural loose end is the third possibility on that list. A secondary mutation or a bypass amplification would still leave him a non-small-cell patient, and every option about to be argued over is a non-small-cell option. Histologic transformation would not: it would mean the disease in his chest has changed into something both proposed regimens treat poorly, and the specimen that could say so is already collected.
Eleven months in, and the fluid nobody's tested yet
Add a SHP2 inhibitor to continued sotorasib. SHP2 sits upstream of most of the pathways tumors use to reactivate RAS-MAPK signaling after G12C inhibition, and in the CodeBreaK 101 phase 1b combination of sotorasib with the SHP2 inhibitor RMC-4630, the investigator-assessed response rate in patients already pretreated with a KRAS G12C inhibitor was 27%. That mechanistic position covers more than one possible resistance route at once.
That 27% comes from a dose-exploration cohort of a handful of patients — the confidence interval around it runs from 6% to 61%, which is another way of saying we don't know the number. And stacking two targeted agents has a real toxicity cost — rash, diarrhea, transaminitis — in a patient already tolerating sotorasib's own side effects only reasonably well. He has had platinum and a checkpoint inhibitor; docetaxel, with or without ramucirumab, is what's actually established after both, and it works regardless of which resistance mechanism is operating.
You're right that the SHP2 mechanism is a plausible fit for reactivation resistance — but plausible mechanism isn't the same as a regimen with real randomized comparative data behind it yet, and I don't think we should ask him to absorb two targeted agents' combined toxicity on the strength of a cohort that small.
You're both arguing about which regimen to give a resistance mechanism neither of you has actually identified. He has a pleural fluid specimen sitting available right now, and the three plausible routes here — a secondary KRAS mutation, bypass amplification, or transformation to small-cell carcinoma — point toward three different next steps, not one. Send it for next-generation sequencing before committing to either the targeted combination or docetaxel exclusively.
A week's real delay against picking the wrong next line for eleven more months isn't a close call. And the transformation question isn't academic here: if that fluid shows small-cell histology, both of the regimens you're arguing about are the wrong answer and he needs platinum-etoposide instead. Start docetaxel now if his symptoms won't wait — it's reasonable no matter what the sequencing shows — and let the SHP2 combination decision wait for an actual answer.
Agreed: pleural fluid sent for resistance next-generation sequencing this week, explicitly including a histology read for small-cell transformation; docetaxel started now for symptom control rather than delaying all treatment while results are pending.
Not agreed: whether SHP2-inhibitor addition should be reserved specifically for a reactivation-consistent sequencing result or considered more broadly once results return regardless of the specific finding — the thoracic oncologist and molecular oncologist differ on how strictly to gate that decision.